Parallel Compression Is a Fast Low-Cost Assay for the High-Throughput Screening of Mechanosensory Cytoskeletal Proteins in Cells

Abstract

Cellular mechanosensing is critical for many biological processes, including cell differentiation, proliferation, migration, and tissue morphogenesis. The actin cytoskeletal proteins play important roles in cellular mechanosensing. Many techniques have been used to investigate the mechanosensory behaviors of these proteins. However, a fast, low-cost assay for the quantitative characterization of these proteins is still lacking. Here, we demonstrate that compression assay using agarose overlay is suitable for the high throughput screening of mechanosensory proteins in live cells while requiring minimal experimental setup. We used several well-studied myosin II mutants to assess the compression assay. On the basis of elasticity theories, we simulated the mechanosensory accumulation of myosin II's and quantitatively reproduced the experimentally observed protein dynamics. Combining the compression assay with confocal microscopy, we monitored the polarization of myosin II oligomers at the subcellular level. The polarization was dependent on the ratio of the two principal strains of the cellular deformations. Finally, we demonstrated that this technique could be used on the investigation of other mechanosensory proteins.

Keywords: actin filament; compression; mechanobiology; mechanosensing; mechanosensory accumulation; mechanotransduction; myosin II.

Figure 1

Deformations of a compressed cell: (a) Schematic graph of the principal strains and the associated coordination; (b) coordination after compression. Distribution of principal strain λ₁ (c) and λ₂ (d) at different compressed states λ₀ = 1.09, 1.16, 1.26, and 1.53 where λ₀ is the deformation at center of the contact region. The top views show the dilation (e) and the tension T₂ (f) for the case of λ₀ = 1.53. The subscripts 1 and 2 represent the meridional and circumferential directions, respectively.

Figure 2

Experimentally observed dynamics of the accumulation of myosin II’s in compressed cells: (a) Time lapse of the accumulation of myosin II 2xELC in multiple cells. (b) Dynamics of the accumulations of WT and 2xELC myosin II’s. (c) Rising phase of the WT myosin II at the lateral edge under different compression forces 2.0, 3.0, and 6.5 nN. (d) Correlation between the myosin II intensity and the circularity of the middle plane of a compressed cell. (e) Correlation between the myosin II intensity and the circumferential strain λ₂. The error bars represent the standard error of mean.

Figure 3

Force-dependent accumulation of myosin II’s observed in experiments. (a) Snapshots of actin probes (GFP-actin and RFP-LimEΔcoil) and myosin II’s (WT, ΔBLCBS, S456L, and 2xELC) were taken by epi-fluorescence microscopy. (b) Compared to actin and LimE, the peak intensities of the accumulation of different myosin II’s displayed distinct trends in the force range of 0 to ~25 nN per cell. The lines only show the trends and are not fitted to the scattered points. The normalized intensity is the ratio between the fluorescence signals at the lateral edge and those in the middle of the cells. Mann–Whitney tests were performed between myosin II’s and actin and the p-values are smaller than 0.01, indicating the myosin II’s behaviors are statistically different from the actin. The error bars represent the standard error of mean.

Figure 4

Spatial distribution of myosin II accumulation. The distribution of GFP-myosin II was imaged by confocal microscopy before (a) and after (b) compression. The upper row is the middle plane, whereas the lower row is the side view of planes indicated by yellow line. (c) The simulated myosin II intensity in compressed cells with different k_area (= 0.5, 1.0,, 1.5 and 2.0) at different compression states λ₀ = 1.16 (left column) and 1.53 (right column). For all cases, D₀ = 1.0 μm²/s and k_d⁰ = 0.5 μM. Scale bars in all panels = 5 μm.

Figure 5

Simulated accumulation of myosin II for different λ₀ (a), α (b), D₀ (c), and k_d⁰ (d). (a), α = 0.0, D₀ = 10.0 μm²/s, and k_d⁰ = 0.5 μM. (b), λ₀ = 1.16, D₀ = 10.0 μm²/s, and k_d⁰ = 0.5 μM. (c), λ₀ = 1.16, α = 0.0, and k_d⁰ = 0.5 μM. (d), λ₀ = 1.16, α = 0.0, and D₀ = 10.0 μm²/s. I_L and I_C represent the concentrations of myosin II in the center and at the lateral side of a compressed cell, respectively.

Figure 6

Polarization of myosin II BTFs due to the anisotropy of principal strains. (a) GFP-myosin II in the middle plane of a compressed cell was imaged by confocal microscopy (upper panel) and the angle distribution of the BTFs was statistically counted (low panel). (b) The fluorescence signals of myosin II were projected onto the middle plane (upper panel), and the angle distribution of the BTFs from other planes (circled by the dotted line) was statistically plotted (lower panel). (c) The polarization of myosin II BTFs was schematically illustrated for different anisotropy of the principal strains. (d) A typical example of the calculated principal strains projected on the middle plane was plotted as a function of the cell radius. The broken line is the boundary between the contact and noncontact regions. (e) The polarization of myosin II BTFs in the contact region was calculated based on the angle deviation ξ = tan⁻¹(βλ₂/λ₁) with β = 4.0. The inserts in (a) and (b) are zoom-in views indicated by the boxes.

Figure 7

Accumulation of other actin cytoskeletal proteins in cells: (a) accumulations of filamin and α-actinin in compressed racE null and myosin II null Dictyostelium cells, respectively. Images were taken by confocal microscopy and combined into stacks (reproduced from Luo et al., permission from Nature Publisher); (b) accumulations of myosin IIB in compressed NIH 3T3 cells. Images were taken by epi-fluorescence microscopy. Scale bars in all panels = 10 μm.